Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Evidence of Nuclear DNA Fragmentation Following Hypoxia‐Ischemia in the Infant Rat Brain, and Transient Forebrain Ischemia in the Adult Gerbil

Evidence of Nuclear DNA Fragmentation Following Hypoxia‐Ischemia in the Infant Rat Brain, and... Servicio Anatornia Patologica, Hospital Principes de Espatia. Universidad de Barcelone, 08907 Hospitalet de Llobregat, Spain 2 Departamento de Chncer y Metastasis, lnstitut de Recerca Oncologica, CSUB, 08907 Barcelona, Spain 3 Unidad de Investigation Biomedica, Hospital Materno-lnfantil. Valle de Hebron, 08038 Barcelona. Spain Department of Neuropathology, Radcliffe Infirmary, Oxford OX2 6HE. U.K. 1 Unidad Neuropatologia, a ladder-type pattern which is typical of nuclear DNA fragmentation into oligonucleosomal fragments (internucleosomal cleavage). These findings suggest that endonuclease(s1 activation may play a role in cell death induced by different forms of hypoxia-ischemia. Introduction Wistar rats, eight days old, were subjected t o permanent bilateral forebrain ischemia, followed by hypoxia for 15 minutes. A cerebral infarct, mainly involving the cerebral neocortex, hippocampus, amygdala, striatum and subcortical white matter was produced. Neurons and glia showing punctate chromatin condensation and karyorrhectic cells were observed 12 hours after hypoxia-ischemia. Their number increased during the first t w o days and recruitment of cells with degenerating nuclei occurred until day five. In situ labeling of nuclear DNA fragmentation stained many normal-appearing nuclei, as well as punctate chromatin condensations and nuclear fragments in karyorrhectic cells. Delayed neuronal death in the C A I area of the hippocampus was observed after 20 minutes of transient forebrain ischemia in the adult gerbil. In situ labeling of nuclear DNA fragmentation demonstrated stained punctate chromatin condensation in a few degenerating cells at 48 hours post-ischemia. Substantial labeling of CAI neurons occurred i n the fourth day. Agarose gel electrophoresis of extracted brain DNA from ischemic infant rats and adult gerbils showed Received: 9 March 1994 Corresponding author: Dr. I. Ferrer, Unidad Neuropatologia. Servicio Anatomia Patologica, Hospital Principes de Esparia, 08907 Hospitalet de Llobregat, Spain Tel. +34 (3) 204 5065; Fax +34 (3) 204 5065 Reduction in the supply of oxygen to the brain may result in cerebral infarction and selective neuronal necrosis (1). Perinatal brain damage is a serious hazard which may produce neurological devastation in human infants (2-4). Although there are many different causes of perinatal brain damage, hypoxicischemic accidents are the origin in a large number of victims. Likewise, transient forebrain ischemia, as observed following cardiac arrest and reperfusion, may have ruinous consequences in different regions, as the cerebral cortex and hippocampus in adult rats. The mechanisms leading t o cell death and tissue necrosis are difficult to analyze in human cases. For this reason, some experimental models of hypoxicischemic injury have been produced in infant rodents i n an attempt t o learn more about factors involved in cell death (5,6). These models reproduce early lesions and late effects of human perinatal brain damage (7). O n t h e other hand, a delayed progressive injury occurs in the CA1 area of the hippocampus in t h e gerbil a few days after transient forebrain ischemia (8-14). Cell death mainly affects pyramidal neurons, whereas gamma aminobutyric acid (GABA)ergic, parvalbumin-immunoreactive neurons of the stratum pyramidale are spared (15,16). Several studies have proposed a calciumdependent excitotoxic mechanism of hypoxic-ischemic cell death in the developing brain on the basis of light and electron microscopical cytopathology, which is similar to that induced by exogenous glutamic acid (17). Extracellular overflow of glutamate and aspartate has also been found in the cortex and basal ganglia of fetal lambs during hypoxia-ischemia (18). Furthermore, treatment with MK-801, a noncompetitive, selective N-methyl-D-aspartate receptor antagonist, reduces striatal and hippocampal lesions I Ferrer et al: Nuclear DNA fragmentation following ischemia pairs (internucleosomal DNA fragmentation), which typically occurs as a result of endonuclease(s) activation, is found after focal or global brain ischemia in adult rats (30-32). The purpose of the present study is to examine the morphology and distribution of dying cells, and to analyze, by in situ specific labeling of nuclear DNA fragmentation and agarose gel electrophoresis of extracted DNA, the possible role of endonuclease(s) activation after hypoxia-ischemia in the infant rat, and following transient forebrain ischemia in the adult gerbil. Material and Methods Figure 1A Normal neocortex in eight days old rats; B Morphological changes in the cerebral neocortex in age-matched animals. 12 hours after hypoxia-ischemia: C Normal hippocampus in eights days old rats; D Morphological changes in the hippocampus in age-matched rats, 12 hours after hypoxiaischemia. Necrosis has extended to all cortical layers. except the molecular layer. Near massive cell death is seen in the dentate gyrus !DG) and hilus (H), whereas patchy necrosis is found in the CA1 region of the hippocampus Hematoxylin and eosin stain. Bar = 50 pm. (19,20), and the calcium antagonist flunarazine limits the extent of the morphologic damage (21). Other studies have also supported the calcium-mediated glutamate toxicity hypothesis t o explain hypoxicischemic cell death, including delayed neuronal death, in adult rats (14,22-25). However, it has been suggested that activation of certain killer proteins may play a role in delayed neuronal death, since protein synthesis inhibitors may reduce the extent of cell death in the ischemic hippocampus (26-29). The possibility that different forms of hypoxiaischemia may produce programmed cell death must be considered since recent observations have suggested that fragmentation of nuclear DNA into oligonucleosomal fragments, multiples of 180-200 base Animals. Wistar rats, eight days old, were anesthetized with halotane (3% for induction, 1.5% for maintenance). Common carotid arteries were exposed through a mid-line anterior neck incision, isolated from the respective vein and nerve, double-ligated with 5-0 surgical silk, and electrocoagulated between t h e two ligatures. The total average time for this surgical procedure was ten minutes. The pups were allowed to recover for 30 to 120 minutes in a warmed cage (35 " C). After this time, groups of 10 t o 12 animals were placed in a airtight 4.0 litre plastic container and placed in a water bath at 37°C. A gas mixture of 7% oxygen, 93% nitrogen was delivered via ventilation for 15 minutes. Thereafter, the pups recovered in room air, in a warmed cage for 30 minutes and then were returned to their own cages. For morphological studies, the rats were sacrificed 1, 12, 24, 48 hours, and 5 or 7 days later, respectively ( n = 3, for every time). This combined procedure (ischemia plus hypoxia) was chosen because very different effects from one animal t o another were produced after ischemia alone. Normal rats and sham-operated animals not subjected t o hypoxia were used as controls. Three t o six months old, adult Mongolian gerbils (Meriones ungiculatus) of both sexes were anaesthetized with halotane mixed with room air, and subjected to 20 minutes of forebrain ischemia by clipping the common carotid arteries as previously reported (16,29). Blood flow during the occlusion, and the reperfusion after the removal of the clips were visually examined. The body temperature was maintained at 36" t o 3 7 ° C during ischemia and the early post-ischemic period. For morphological studies, gerbils were sacrificed at 48 hours ( n = 4), or four days (n = 10) after ischemia. Age-matched gerbils and sham-operated animals were used as controls. Morphological studies and in situ labeling of nuclear DNA fragmentation. The animals were reanesthetized with deep diethyl-ether and perfused through the heart with 4% paraformaldehyde in phosphate buffer. The brains were rapidly removed and fixed in a similar solution for 24 hours. After this time, the tissue was embedded in paraffin and I. Ferrer et al: Nuclear DNA fragmentation following ischemia serial-cut 5 pm thick coronal sections with a sliding microtome, and collected. Dewaxed sections were stained with hematoxylin and eosin, or processed for in situ labeling of nuclear DNA fragmentation (33). This method is based on the incorporation of biotin16-2'-deoxy-uridine-5'triphosphate (biotin-16-dUTP) at sites of DNA break by means of a DNA deoxynucleotidylexotransferase (terminal transferase TdT). The signal is amplified by avidin-peroxidase (ABC kit, Vector Labs), enabling identification by light microscopy. The protocol used was originally described by Gavrieli et al. (33), with the following modifications: Dewaxed sections were treated with 20 p g / m l proteinase K for 15 minutes and then with H202 for 5 minutes. After washing with TdT buffer for 15 minutes, the tissue sections were then incubated with 28 pl dUTP, 12 pl TdT in 1000 pl of TdT buffer for 60 minutes at 37°C. Thereafter, the sections were washed with TB buffer (300 mM NaC12 and 30 mM sodium citrate) for 15 minutes, 2% bovine albumin for 10 minutes, phosphate buffer for 5 minutes, and incubated with ABC at a dilution of 1:25 for 30 minutes at 37" C. The immunoreaction was visualized with 0.05% diaminobenzidine and 0.0196 H202. The TdT buffer was composed of 200 mM cacodylic acid, 200 mM KCl and 25 mM TRlS or trizma base at pH 6.6. Bovine albumin (1.25 m g / m l ) and cobalt chloride (40 pllml) were added to TdT buffer before use. Biotin-16-dUTP and TdT were obtained from Boehringer Mannheim. Figure 2A Normal dentate neurons of a eight day old rat; B Dying cells in the dentate gyrus of a eight day old rat, 12 hours after hypoxia-ischemia, show extreme chromatin condensation and nuclear fragmentation. Hematoxylin and eosin stain. Bar = 25 pm. were exposed at - 8 0 ° C t o X-ray film with intensifying screens. Results Agarose gel electrophoresis of extracted DNA. For biochemical studies, the brains of hypoxic-ischemic rats obtained 1 hour or 12 hours after t h e insult, and dissected hippocampi of gerbils at day four post-ischemia, as well as the corresponding controls ( n = 4 , for every group), were obtained fresh and rapidly frozen in liquid nitrogen, then stored at 80°C until use. About 80 mg of tissue were homogenized in 2 ml lysis buffer (ASAP kit, Boehringer Mannheim), using a hand-operated homogenizer. The homogenate was transferred t o an eppendorf tube, and incubated with 60 p1 RNAase (10 mglml) at 37°C for one hour, followed by incubation with 100 pl proteinase K (20 m g / m l ) and 215 p1 5 M guanidine HCl, at 60"C, for three hours. Equal amounts of DNA (approx. 10 pg) were run in each lane of a 1.5% agarose gel containing ethidium bromide (0.5 ug/ml), and electrophoresed in TAE buffer at 100 mV for 45-60 minutes. Gels included molecular weight markers and positive controls for internucleosomal DNA fragmentation (liver of newborn rats subjected to hypoxia). In order to improve the detection of DNA fragments, DNA was denatured and transferred to nylon membranes (Hybond-N, Amersham, Arlington Heights, IL, U.S.A.), according t o the technique of Southern blotting, and then hybridized with a radiolabeled probe (34). Filters Hypoxia-ischemia in the infant rat. Tissue necrosis was observed 12 hours after the hypoxic-ischemic insult in the infant rats. The areas mainly affected were the cerebral neocortex, including the interhemispheric cortex, hippocampus, amygdala, striatum and subcortical white matter. All of the cortical layers, except for the molecular layer, were involved (Figs. lA,lB). A columnar pattern of cellular damage was seen in one rat. The dentate gyrus was severely damaged, whereas the hippocampus proper showed patchy necrosis of the CAI region; both the CA2 and CA3 regions were less damaged (Figs. lC,lD). Lesions extended 24 hours after injury to the habenular complex and entorhinal cortex. The cerebellum and the brainstem were spared. Massive neuronal loss occurred by day five, and cavitation of the cerebral cortex and striatum by day seven. Cellular shrinkage of large pyramidal neurons, and swelling of the small cortical and striatal neurons, together with spongiosis, predominated in the first 24 hours. Karyorrhectic cells, and cells with punctate chromatin condensation, were seen as early as 12 hours after hypoxia-ischemia in the cerebral cortex, subcortical white matter, and hippocampus proper. Neurons with fragmented nuclei prevailed I. Ferrer et al: Nuclear DNA fragmentation following ischemia matin condensation in other cells. Large nuclear globules and protrusions occurred in some pyramidal neurons of the cerebral cortex and CA1 region of the hippocampus. Finally, irregular nuclear fragments were seen in most damaged small and medium-sized cells (Fig. 4 . ) Transient forebrain ischemia in the adult gerbil. The second day after ischemia, t h e cytoplasm of most neurons in the CA1 area was shrunken but few neurons had punctate chromatin condensation. On day four, the cytoplasm of the majority of CAI neurons was swollen, and the nuclei of most neurons exhibited punctate condensation of the chromatin (Fig. 5). In situ labeling of nuclear DNA fragmentation revealed a few CAI pyramidal cells 48 hours post-ischemia (data not shown). Labeling extended throughout the CA1 area at day four post-ischemia (Figs. 6A,6B). Punctate chromatin condensation was clearly decorated by the peroxidase reaction (Fig. 6C). Agarose gel electrophoresis. Agarose gel electrophoresis of extracted DNA was carried out in brain samples obtained 1 hour or 12 hours after the insult in infant rats. No anomalies were observed during the first hour, but a typical ladder-type pattern was disclosed 12 hours after hypoxia-ischemia by using Southern hybridization with 32P-labeled rat genomic DNA (Fig. 7). Ethidium bromide staining of agarose gel revealed a discrete ladder of oligonucleosomal fragments in samples of DNA extracted from ischemic hippocampus in one out of four experimental animals. Southern hybridization with 32P-labeled rat genomic DNA clearly disclosed laddering in the hippocampus of the four experimental animals (Fig. 8). Discussion Figure 3A Karyorrhectic cells (KC) and cells w i t h punctate chromatin condensation (arrowhead1 in the cerebral neocortex; 8 KC and cells with punctate chromatin condensation (arrowhead) in the amygdala: C KC in the subcortical white matter; D KC in the CAI region of the hippocampus, in eight days old rats, 12 hours after hypoxia-ischemia; E KC in the hippocampus, F KC in the cerebral neocortex. G KC in the amygdala, in rats subjected t o hypoxia-ischemia at the age of eight days and sacrificed 48 hours later. Hematoxylin and eosin stain. Bar = 10 pm. in the dentate gyrus during the same period (Fig. 2). Cells with punctate chromatin condensation and karyorrhectic cells increased in number by day two, and, although scattered in different brain regions, they dominated in the upper cortical layers, hippocampus and amygdala (Fig. 3 ) . On the basis of their morphology and distribution, cells with fragmented nuclei included neurons, glial cells and endothelial cells. Macrophages augmented in damaged and neighboring areas from day two onwards. Reactive astrocytes were seen by day five. Although lessening in number, cells with punctate chromatin condensation and karyorrhectic cells were recruited until day five. In sitzr labeling of nuclear DNA fragmentation showed stained cells 12 hours after the insult. Their number dramatically increased during the following hours. The morphology of stained cells was variable. Some stained nuclei exhibited normal morphology, whereas the peroxidase reaction decorated fine chro- The main damaged areas in infant rats subjected to hypoxia-ischemia are the cerebral neocortex, hippocampus, striatum, a mygdala and subcortical white matter. In the neocortex, lesions affect all cortical layers, whereas in the hippocampal complex, the dentate gyrus is the most severely devastated, followed by the CA1 region. As early as 12 hours after the hypoxic-ischemic insult, many cells with punctate chromatin condensation, and cells with fragmented nuclei (karyorrhectic cells) are recognized. The number of these cells increases in the following two days, and decreases thereafter so that few are observed by the end of the first week. Recruitment of karyorrhectic cells in the upper cortical layers, subcortical white matter, dentate gyrus, amygdala, and striatum occurs until day five. These findings indicate that hypoxia-ischemia in developing rats induce different types of nuclear anomalies, including karyorrhexis in both neurons and glial cells. Karyorrhectic cells are also found in human infants after prenatal brain I. Ferrer et al: Nuclear DNA fragmentation following ischemia injury (3,35). Delayed neuronal death in the CAI area of the hippocampus in the gerbil after transient forebrain ischemia is characterized by primary degeneration of the cytoplasm, followed by punctate chromatin condensation. Early involvement of the cytoplasm is demonstrated by direct ultrastructural observation of abnormal organelles, together with reduced MAP-2 immunoreactivity, and early destruction of tubulin, microtubules and neurofilaments (10,13,14,29,31-41). In situ labeling of nuclear DNA fragmentation (33) stains most, if not all, karyorrhectic cells, in addition to cells with punctate chromatin condensation, and many normal-looking nuclei in developing rats subjected to hypoxia-ischemia. This method also has shown that DNA fragmentation first occurs in some CA1 hippocampal neurons, 48 hours after 20 minutes of transient forebrain ischemia, and that DNA fragmentation is almost massive at day four postischemia in the hippocampus of the gerbil. Therefore, although the method of in situ labeling of nuclear DNA fragmentation serves t o recognize apoptotic cells as originally proposed (33), the present findings indicate that not only apoptotic cells, but other degenerating nuclei bearing fragmented DNA can be recognized as well. Agarose gel electrophoresis of extracted DNA from hypoxic-ischemic brains of infant rats, and from t h e hippocampus of adult gerbils subjected t o transient forebrain ischemia, has shown a characteristic ladder pattern which recognizes regular fragments of DNA multiples of approximately 180-200 base pairs. This ladder pattern is typical of endonuclease-mediated, doublestrand cleavage of nuclear DNA in apoptosis (42-48). Internucleosomal cleavage has also been observed in different conditions with prevailing necrosis, if apoptosis is also present (49,50). Apoptosis is classically defined by the combination of typical morphological and biochemical criteria (51). However, different lines of cultured epithelial cells undergoing apoptosis may show cleavage of DNA to 300 and/ or 50 kilobase fragments prior to or in the absence of oligonucleosomal fragments resulting from endonuclease activation (52). Therefore, these observations suggest that endonuclease activation is not mandatory nor exclusive of apoptosis, but rather that it may be a common feature of many types of cell death. Endonuclease activation has also been found in necrosis following focal ischemic injury in the adult rat brain (31,32) and focal cortical freeze injury of rats ( 5 3 ) . DNA fragmentation in the last two rat models occurs as early as 3 hours after injury, reaching a maximum at 24 hours, and then declines. This situation resembles DNA fragmentation following hypoxia-ischemia in t h e infant rat, although recruitment of dying cells occurs in the last experimental model, but is in contrast with delayed postischemic neuronal death which starts 48 hours after Figure 4 In sifu labeling of nuclear DNA fragmentation in eight days old rats subjected to hypoxia-ischemia and sacrificed 12 hours later. A cerebral neocortex; B CA1 region of the hippocampus. Karyorrhectic cells, as well as cells with punctate chromatin condensation (arrow heads) and apparently normal cells (arrows) are observed. Sections without counterstaining. Bar = 10 pm. injury and reaches peak values by day four. Taken together, the present morphological and biochemical findings are in accordance with the rationale that different types of cell death may be produced after different hypoxic-ischemic insults in the developing Figure 5A CA1 area of the hippocampus of a sham-operated control adult gerbil; B CA1 area of the gerbil hippocampus, four days after 20 minutes of bilateral forebrain ischemia showing dying cells with punctate chromatin condensation Hematoxylin and eosin stain. Bar = 100 pm. I. Ferrer et al: Nuclear DNA fragmentation following ischemia and adult brain. The technique of in situ labeling of nuclear DNA fragmentation is a useful tool t o recognize DNA fragmentation even in nuclei which appear normal in conventional stains, as well as in cells with punctate chromatin condensation and karyorrhectic cells, in addition to apoptotic cells (33). Internucleosomal DNA fragmentation, as revealed with agarose gel electrophoresis, strongly suggests that hypoxia-ischemia may kill an indeterminate number of cells in association with a process which may be mediated by endonuclease(s). However, further studies are necessary to elucidate the localization and characteristics of endonuclease(s) activated in different forms of post-ischemic cell death. Acknowledgements The research presented was supported by the Fondo de lnvestigacion Sanitaria and grants from the Pi i Sunyer Foundation, Ministerio de Educacion y Ciencia and Comissio Interdepartamental de Recerca i Innovaci6 Tecnologica. The authors would like t o thank Mr. T. Yohannan for editorial assistance. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Brain Pathology Wiley

Evidence of Nuclear DNA Fragmentation Following Hypoxia‐Ischemia in the Infant Rat Brain, and Transient Forebrain Ischemia in the Adult Gerbil

Loading next page...
 
/lp/wiley/evidence-of-nuclear-dna-fragmentation-following-hypoxia-ischemia-in-lhQBn802Dy

References (53)

Publisher
Wiley
Copyright
Copyright © 1994 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1015-6305
eISSN
1750-3639
DOI
10.1111/j.1750-3639.1994.tb00821.x
Publisher site
See Article on Publisher Site

Abstract

Servicio Anatornia Patologica, Hospital Principes de Espatia. Universidad de Barcelone, 08907 Hospitalet de Llobregat, Spain 2 Departamento de Chncer y Metastasis, lnstitut de Recerca Oncologica, CSUB, 08907 Barcelona, Spain 3 Unidad de Investigation Biomedica, Hospital Materno-lnfantil. Valle de Hebron, 08038 Barcelona. Spain Department of Neuropathology, Radcliffe Infirmary, Oxford OX2 6HE. U.K. 1 Unidad Neuropatologia, a ladder-type pattern which is typical of nuclear DNA fragmentation into oligonucleosomal fragments (internucleosomal cleavage). These findings suggest that endonuclease(s1 activation may play a role in cell death induced by different forms of hypoxia-ischemia. Introduction Wistar rats, eight days old, were subjected t o permanent bilateral forebrain ischemia, followed by hypoxia for 15 minutes. A cerebral infarct, mainly involving the cerebral neocortex, hippocampus, amygdala, striatum and subcortical white matter was produced. Neurons and glia showing punctate chromatin condensation and karyorrhectic cells were observed 12 hours after hypoxia-ischemia. Their number increased during the first t w o days and recruitment of cells with degenerating nuclei occurred until day five. In situ labeling of nuclear DNA fragmentation stained many normal-appearing nuclei, as well as punctate chromatin condensations and nuclear fragments in karyorrhectic cells. Delayed neuronal death in the C A I area of the hippocampus was observed after 20 minutes of transient forebrain ischemia in the adult gerbil. In situ labeling of nuclear DNA fragmentation demonstrated stained punctate chromatin condensation in a few degenerating cells at 48 hours post-ischemia. Substantial labeling of CAI neurons occurred i n the fourth day. Agarose gel electrophoresis of extracted brain DNA from ischemic infant rats and adult gerbils showed Received: 9 March 1994 Corresponding author: Dr. I. Ferrer, Unidad Neuropatologia. Servicio Anatomia Patologica, Hospital Principes de Esparia, 08907 Hospitalet de Llobregat, Spain Tel. +34 (3) 204 5065; Fax +34 (3) 204 5065 Reduction in the supply of oxygen to the brain may result in cerebral infarction and selective neuronal necrosis (1). Perinatal brain damage is a serious hazard which may produce neurological devastation in human infants (2-4). Although there are many different causes of perinatal brain damage, hypoxicischemic accidents are the origin in a large number of victims. Likewise, transient forebrain ischemia, as observed following cardiac arrest and reperfusion, may have ruinous consequences in different regions, as the cerebral cortex and hippocampus in adult rats. The mechanisms leading t o cell death and tissue necrosis are difficult to analyze in human cases. For this reason, some experimental models of hypoxicischemic injury have been produced in infant rodents i n an attempt t o learn more about factors involved in cell death (5,6). These models reproduce early lesions and late effects of human perinatal brain damage (7). O n t h e other hand, a delayed progressive injury occurs in the CA1 area of the hippocampus in t h e gerbil a few days after transient forebrain ischemia (8-14). Cell death mainly affects pyramidal neurons, whereas gamma aminobutyric acid (GABA)ergic, parvalbumin-immunoreactive neurons of the stratum pyramidale are spared (15,16). Several studies have proposed a calciumdependent excitotoxic mechanism of hypoxic-ischemic cell death in the developing brain on the basis of light and electron microscopical cytopathology, which is similar to that induced by exogenous glutamic acid (17). Extracellular overflow of glutamate and aspartate has also been found in the cortex and basal ganglia of fetal lambs during hypoxia-ischemia (18). Furthermore, treatment with MK-801, a noncompetitive, selective N-methyl-D-aspartate receptor antagonist, reduces striatal and hippocampal lesions I Ferrer et al: Nuclear DNA fragmentation following ischemia pairs (internucleosomal DNA fragmentation), which typically occurs as a result of endonuclease(s) activation, is found after focal or global brain ischemia in adult rats (30-32). The purpose of the present study is to examine the morphology and distribution of dying cells, and to analyze, by in situ specific labeling of nuclear DNA fragmentation and agarose gel electrophoresis of extracted DNA, the possible role of endonuclease(s) activation after hypoxia-ischemia in the infant rat, and following transient forebrain ischemia in the adult gerbil. Material and Methods Figure 1A Normal neocortex in eight days old rats; B Morphological changes in the cerebral neocortex in age-matched animals. 12 hours after hypoxia-ischemia: C Normal hippocampus in eights days old rats; D Morphological changes in the hippocampus in age-matched rats, 12 hours after hypoxiaischemia. Necrosis has extended to all cortical layers. except the molecular layer. Near massive cell death is seen in the dentate gyrus !DG) and hilus (H), whereas patchy necrosis is found in the CA1 region of the hippocampus Hematoxylin and eosin stain. Bar = 50 pm. (19,20), and the calcium antagonist flunarazine limits the extent of the morphologic damage (21). Other studies have also supported the calcium-mediated glutamate toxicity hypothesis t o explain hypoxicischemic cell death, including delayed neuronal death, in adult rats (14,22-25). However, it has been suggested that activation of certain killer proteins may play a role in delayed neuronal death, since protein synthesis inhibitors may reduce the extent of cell death in the ischemic hippocampus (26-29). The possibility that different forms of hypoxiaischemia may produce programmed cell death must be considered since recent observations have suggested that fragmentation of nuclear DNA into oligonucleosomal fragments, multiples of 180-200 base Animals. Wistar rats, eight days old, were anesthetized with halotane (3% for induction, 1.5% for maintenance). Common carotid arteries were exposed through a mid-line anterior neck incision, isolated from the respective vein and nerve, double-ligated with 5-0 surgical silk, and electrocoagulated between t h e two ligatures. The total average time for this surgical procedure was ten minutes. The pups were allowed to recover for 30 to 120 minutes in a warmed cage (35 " C). After this time, groups of 10 t o 12 animals were placed in a airtight 4.0 litre plastic container and placed in a water bath at 37°C. A gas mixture of 7% oxygen, 93% nitrogen was delivered via ventilation for 15 minutes. Thereafter, the pups recovered in room air, in a warmed cage for 30 minutes and then were returned to their own cages. For morphological studies, the rats were sacrificed 1, 12, 24, 48 hours, and 5 or 7 days later, respectively ( n = 3, for every time). This combined procedure (ischemia plus hypoxia) was chosen because very different effects from one animal t o another were produced after ischemia alone. Normal rats and sham-operated animals not subjected t o hypoxia were used as controls. Three t o six months old, adult Mongolian gerbils (Meriones ungiculatus) of both sexes were anaesthetized with halotane mixed with room air, and subjected to 20 minutes of forebrain ischemia by clipping the common carotid arteries as previously reported (16,29). Blood flow during the occlusion, and the reperfusion after the removal of the clips were visually examined. The body temperature was maintained at 36" t o 3 7 ° C during ischemia and the early post-ischemic period. For morphological studies, gerbils were sacrificed at 48 hours ( n = 4), or four days (n = 10) after ischemia. Age-matched gerbils and sham-operated animals were used as controls. Morphological studies and in situ labeling of nuclear DNA fragmentation. The animals were reanesthetized with deep diethyl-ether and perfused through the heart with 4% paraformaldehyde in phosphate buffer. The brains were rapidly removed and fixed in a similar solution for 24 hours. After this time, the tissue was embedded in paraffin and I. Ferrer et al: Nuclear DNA fragmentation following ischemia serial-cut 5 pm thick coronal sections with a sliding microtome, and collected. Dewaxed sections were stained with hematoxylin and eosin, or processed for in situ labeling of nuclear DNA fragmentation (33). This method is based on the incorporation of biotin16-2'-deoxy-uridine-5'triphosphate (biotin-16-dUTP) at sites of DNA break by means of a DNA deoxynucleotidylexotransferase (terminal transferase TdT). The signal is amplified by avidin-peroxidase (ABC kit, Vector Labs), enabling identification by light microscopy. The protocol used was originally described by Gavrieli et al. (33), with the following modifications: Dewaxed sections were treated with 20 p g / m l proteinase K for 15 minutes and then with H202 for 5 minutes. After washing with TdT buffer for 15 minutes, the tissue sections were then incubated with 28 pl dUTP, 12 pl TdT in 1000 pl of TdT buffer for 60 minutes at 37°C. Thereafter, the sections were washed with TB buffer (300 mM NaC12 and 30 mM sodium citrate) for 15 minutes, 2% bovine albumin for 10 minutes, phosphate buffer for 5 minutes, and incubated with ABC at a dilution of 1:25 for 30 minutes at 37" C. The immunoreaction was visualized with 0.05% diaminobenzidine and 0.0196 H202. The TdT buffer was composed of 200 mM cacodylic acid, 200 mM KCl and 25 mM TRlS or trizma base at pH 6.6. Bovine albumin (1.25 m g / m l ) and cobalt chloride (40 pllml) were added to TdT buffer before use. Biotin-16-dUTP and TdT were obtained from Boehringer Mannheim. Figure 2A Normal dentate neurons of a eight day old rat; B Dying cells in the dentate gyrus of a eight day old rat, 12 hours after hypoxia-ischemia, show extreme chromatin condensation and nuclear fragmentation. Hematoxylin and eosin stain. Bar = 25 pm. were exposed at - 8 0 ° C t o X-ray film with intensifying screens. Results Agarose gel electrophoresis of extracted DNA. For biochemical studies, the brains of hypoxic-ischemic rats obtained 1 hour or 12 hours after t h e insult, and dissected hippocampi of gerbils at day four post-ischemia, as well as the corresponding controls ( n = 4 , for every group), were obtained fresh and rapidly frozen in liquid nitrogen, then stored at 80°C until use. About 80 mg of tissue were homogenized in 2 ml lysis buffer (ASAP kit, Boehringer Mannheim), using a hand-operated homogenizer. The homogenate was transferred t o an eppendorf tube, and incubated with 60 p1 RNAase (10 mglml) at 37°C for one hour, followed by incubation with 100 pl proteinase K (20 m g / m l ) and 215 p1 5 M guanidine HCl, at 60"C, for three hours. Equal amounts of DNA (approx. 10 pg) were run in each lane of a 1.5% agarose gel containing ethidium bromide (0.5 ug/ml), and electrophoresed in TAE buffer at 100 mV for 45-60 minutes. Gels included molecular weight markers and positive controls for internucleosomal DNA fragmentation (liver of newborn rats subjected to hypoxia). In order to improve the detection of DNA fragments, DNA was denatured and transferred to nylon membranes (Hybond-N, Amersham, Arlington Heights, IL, U.S.A.), according t o the technique of Southern blotting, and then hybridized with a radiolabeled probe (34). Filters Hypoxia-ischemia in the infant rat. Tissue necrosis was observed 12 hours after the hypoxic-ischemic insult in the infant rats. The areas mainly affected were the cerebral neocortex, including the interhemispheric cortex, hippocampus, amygdala, striatum and subcortical white matter. All of the cortical layers, except for the molecular layer, were involved (Figs. lA,lB). A columnar pattern of cellular damage was seen in one rat. The dentate gyrus was severely damaged, whereas the hippocampus proper showed patchy necrosis of the CAI region; both the CA2 and CA3 regions were less damaged (Figs. lC,lD). Lesions extended 24 hours after injury to the habenular complex and entorhinal cortex. The cerebellum and the brainstem were spared. Massive neuronal loss occurred by day five, and cavitation of the cerebral cortex and striatum by day seven. Cellular shrinkage of large pyramidal neurons, and swelling of the small cortical and striatal neurons, together with spongiosis, predominated in the first 24 hours. Karyorrhectic cells, and cells with punctate chromatin condensation, were seen as early as 12 hours after hypoxia-ischemia in the cerebral cortex, subcortical white matter, and hippocampus proper. Neurons with fragmented nuclei prevailed I. Ferrer et al: Nuclear DNA fragmentation following ischemia matin condensation in other cells. Large nuclear globules and protrusions occurred in some pyramidal neurons of the cerebral cortex and CA1 region of the hippocampus. Finally, irregular nuclear fragments were seen in most damaged small and medium-sized cells (Fig. 4 . ) Transient forebrain ischemia in the adult gerbil. The second day after ischemia, t h e cytoplasm of most neurons in the CA1 area was shrunken but few neurons had punctate chromatin condensation. On day four, the cytoplasm of the majority of CAI neurons was swollen, and the nuclei of most neurons exhibited punctate condensation of the chromatin (Fig. 5). In situ labeling of nuclear DNA fragmentation revealed a few CAI pyramidal cells 48 hours post-ischemia (data not shown). Labeling extended throughout the CA1 area at day four post-ischemia (Figs. 6A,6B). Punctate chromatin condensation was clearly decorated by the peroxidase reaction (Fig. 6C). Agarose gel electrophoresis. Agarose gel electrophoresis of extracted DNA was carried out in brain samples obtained 1 hour or 12 hours after the insult in infant rats. No anomalies were observed during the first hour, but a typical ladder-type pattern was disclosed 12 hours after hypoxia-ischemia by using Southern hybridization with 32P-labeled rat genomic DNA (Fig. 7). Ethidium bromide staining of agarose gel revealed a discrete ladder of oligonucleosomal fragments in samples of DNA extracted from ischemic hippocampus in one out of four experimental animals. Southern hybridization with 32P-labeled rat genomic DNA clearly disclosed laddering in the hippocampus of the four experimental animals (Fig. 8). Discussion Figure 3A Karyorrhectic cells (KC) and cells w i t h punctate chromatin condensation (arrowhead1 in the cerebral neocortex; 8 KC and cells with punctate chromatin condensation (arrowhead) in the amygdala: C KC in the subcortical white matter; D KC in the CAI region of the hippocampus, in eight days old rats, 12 hours after hypoxia-ischemia; E KC in the hippocampus, F KC in the cerebral neocortex. G KC in the amygdala, in rats subjected t o hypoxia-ischemia at the age of eight days and sacrificed 48 hours later. Hematoxylin and eosin stain. Bar = 10 pm. in the dentate gyrus during the same period (Fig. 2). Cells with punctate chromatin condensation and karyorrhectic cells increased in number by day two, and, although scattered in different brain regions, they dominated in the upper cortical layers, hippocampus and amygdala (Fig. 3 ) . On the basis of their morphology and distribution, cells with fragmented nuclei included neurons, glial cells and endothelial cells. Macrophages augmented in damaged and neighboring areas from day two onwards. Reactive astrocytes were seen by day five. Although lessening in number, cells with punctate chromatin condensation and karyorrhectic cells were recruited until day five. In sitzr labeling of nuclear DNA fragmentation showed stained cells 12 hours after the insult. Their number dramatically increased during the following hours. The morphology of stained cells was variable. Some stained nuclei exhibited normal morphology, whereas the peroxidase reaction decorated fine chro- The main damaged areas in infant rats subjected to hypoxia-ischemia are the cerebral neocortex, hippocampus, striatum, a mygdala and subcortical white matter. In the neocortex, lesions affect all cortical layers, whereas in the hippocampal complex, the dentate gyrus is the most severely devastated, followed by the CA1 region. As early as 12 hours after the hypoxic-ischemic insult, many cells with punctate chromatin condensation, and cells with fragmented nuclei (karyorrhectic cells) are recognized. The number of these cells increases in the following two days, and decreases thereafter so that few are observed by the end of the first week. Recruitment of karyorrhectic cells in the upper cortical layers, subcortical white matter, dentate gyrus, amygdala, and striatum occurs until day five. These findings indicate that hypoxia-ischemia in developing rats induce different types of nuclear anomalies, including karyorrhexis in both neurons and glial cells. Karyorrhectic cells are also found in human infants after prenatal brain I. Ferrer et al: Nuclear DNA fragmentation following ischemia injury (3,35). Delayed neuronal death in the CAI area of the hippocampus in the gerbil after transient forebrain ischemia is characterized by primary degeneration of the cytoplasm, followed by punctate chromatin condensation. Early involvement of the cytoplasm is demonstrated by direct ultrastructural observation of abnormal organelles, together with reduced MAP-2 immunoreactivity, and early destruction of tubulin, microtubules and neurofilaments (10,13,14,29,31-41). In situ labeling of nuclear DNA fragmentation (33) stains most, if not all, karyorrhectic cells, in addition to cells with punctate chromatin condensation, and many normal-looking nuclei in developing rats subjected to hypoxia-ischemia. This method also has shown that DNA fragmentation first occurs in some CA1 hippocampal neurons, 48 hours after 20 minutes of transient forebrain ischemia, and that DNA fragmentation is almost massive at day four postischemia in the hippocampus of the gerbil. Therefore, although the method of in situ labeling of nuclear DNA fragmentation serves t o recognize apoptotic cells as originally proposed (33), the present findings indicate that not only apoptotic cells, but other degenerating nuclei bearing fragmented DNA can be recognized as well. Agarose gel electrophoresis of extracted DNA from hypoxic-ischemic brains of infant rats, and from t h e hippocampus of adult gerbils subjected t o transient forebrain ischemia, has shown a characteristic ladder pattern which recognizes regular fragments of DNA multiples of approximately 180-200 base pairs. This ladder pattern is typical of endonuclease-mediated, doublestrand cleavage of nuclear DNA in apoptosis (42-48). Internucleosomal cleavage has also been observed in different conditions with prevailing necrosis, if apoptosis is also present (49,50). Apoptosis is classically defined by the combination of typical morphological and biochemical criteria (51). However, different lines of cultured epithelial cells undergoing apoptosis may show cleavage of DNA to 300 and/ or 50 kilobase fragments prior to or in the absence of oligonucleosomal fragments resulting from endonuclease activation (52). Therefore, these observations suggest that endonuclease activation is not mandatory nor exclusive of apoptosis, but rather that it may be a common feature of many types of cell death. Endonuclease activation has also been found in necrosis following focal ischemic injury in the adult rat brain (31,32) and focal cortical freeze injury of rats ( 5 3 ) . DNA fragmentation in the last two rat models occurs as early as 3 hours after injury, reaching a maximum at 24 hours, and then declines. This situation resembles DNA fragmentation following hypoxia-ischemia in t h e infant rat, although recruitment of dying cells occurs in the last experimental model, but is in contrast with delayed postischemic neuronal death which starts 48 hours after Figure 4 In sifu labeling of nuclear DNA fragmentation in eight days old rats subjected to hypoxia-ischemia and sacrificed 12 hours later. A cerebral neocortex; B CA1 region of the hippocampus. Karyorrhectic cells, as well as cells with punctate chromatin condensation (arrow heads) and apparently normal cells (arrows) are observed. Sections without counterstaining. Bar = 10 pm. injury and reaches peak values by day four. Taken together, the present morphological and biochemical findings are in accordance with the rationale that different types of cell death may be produced after different hypoxic-ischemic insults in the developing Figure 5A CA1 area of the hippocampus of a sham-operated control adult gerbil; B CA1 area of the gerbil hippocampus, four days after 20 minutes of bilateral forebrain ischemia showing dying cells with punctate chromatin condensation Hematoxylin and eosin stain. Bar = 100 pm. I. Ferrer et al: Nuclear DNA fragmentation following ischemia and adult brain. The technique of in situ labeling of nuclear DNA fragmentation is a useful tool t o recognize DNA fragmentation even in nuclei which appear normal in conventional stains, as well as in cells with punctate chromatin condensation and karyorrhectic cells, in addition to apoptotic cells (33). Internucleosomal DNA fragmentation, as revealed with agarose gel electrophoresis, strongly suggests that hypoxia-ischemia may kill an indeterminate number of cells in association with a process which may be mediated by endonuclease(s). However, further studies are necessary to elucidate the localization and characteristics of endonuclease(s) activated in different forms of post-ischemic cell death. Acknowledgements The research presented was supported by the Fondo de lnvestigacion Sanitaria and grants from the Pi i Sunyer Foundation, Ministerio de Educacion y Ciencia and Comissio Interdepartamental de Recerca i Innovaci6 Tecnologica. The authors would like t o thank Mr. T. Yohannan for editorial assistance.

Journal

Brain PathologyWiley

Published: Apr 1, 1994

There are no references for this article.